During our continued effort to measure iron level in tissue, we have developed two novel methods for the measurement of magnetic susceptibility. Before applying these techniques to humans, we need to conduct thorough phantom and animal studies. Significant progress has been made in this direction. In the first technique, called the contact reference method, a series of gerbil mice were injected iron dextran solutions to induce iron overload to different levels. Then we measured their liver and heart magnetic susceptibilities ex vivo. At the same time, these samples were sent to an independent lab to measure their iron concentrations using atomic absorption spectroscopy. We have found that sample magnetic susceptibilities are linearly proportional to their iron concentrations. Another susceptibility measurement we have developed is the dipole field model. First, a test tube was inserted in a big flask filled with water. We acquired the phase image of a slice in the coronal direction. Then a small sample (solid or liquid) was placed in the test tube. Under the same experimental conditions, we took the measurement again. The two phase image difference provides the magnetic field generated by the sample magnetization. We approximated the sample to a magnetic dipole, ignoring all other higher order field, and then modelled the magnetic field of sample to the dipole field formula. This way, we could extract the magnetic susceptibility of sample. Computer simulation shows that the dipole field approximation introduces very small error. Currently we are seeking to apply this method to mapping brain magnetic susceptibility.